This panel study has two specific aims: 1) to provide information regarding the effect of race and gender on specific psychosocial responses (grief response, self-esteem, and relationship adjustment) to perinatal loss over time, and 2) to provide information regarding the impact of certain subject attribute variables (parental age, marital status, income level, pregnancy planning status and parity status) on specific psychosocial responses (grief response, self-esteem, and relationship adjustment) to perinatal loss over time. The purpose of the study is to extend prior research on the impact of perinatal loss events on parents by comparing over time, the grief response, self-esteem and relationship adjustment of white and African-American parents. The long term goal of this study is to identify nursing interventions to aid persons at risk for psychosocial difficulties related to the impact of perinatal loss. A sample of 200 couples will be recruited into the study: 100 white and 100 African-American for a total of 400 subjects. Four physician obstetrical and gynecological practices will be used to access the sample. Two of the physicians have caseloads of primarily African-American patients; two physicians have primarily white patient caseloads. Sample selection will be limited to those patients who have a physician generated diagnosis related to perinatal loss. Approximately 3 months post diagnosis, a mailed survey will be sent to the homes of the parents. Parents will be asked to complete the Perinatal Grief Scale (33 items), the Rosenberg Self-esteem Scale (10 items) and the Locke-Wallace Relationship Inventory (16 items). Demographic information will also be collected. Three months after the initial contact, parents will again be contacted to complete the inventories. A repeated measure multivariate analysis of covariance (MANCOVA) and regression analysis will be used to address the research questions. The research hypotheses will be tested using Pillai's Criterion at a significance level of .05. Exploratory analysis will be done to investigate the fit of the data regarding the underlying assumptions necessary for use of the MANCOVA technique.